Create app.py

app.py

@@ -0,0 +1,113 @@
+import os
+import pdfplumber
+from dotenv import load_dotenv
+import gradio as gr
+from langchain_openai import OpenAIEmbeddings, ChatOpenAI
+from langchain_chroma import Chroma
+from langchain.text_splitter import RecursiveCharacterTextSplitter
+from langchain.chains import RetrievalQA
+from langchain.llms import OpenAI
+from langchain.document_loaders import TextLoader
+from langchain.docstore.document import Document
+from transformers import AutoTokenizer
+from langchain.document_loaders import PyPDFLoader
+from langchain.memory import ConversationBufferMemory
+from langchain.chains import ConversationalRetrievalChain
+
+
+
+# price is a factor for our company, so we're going to use a low cost model
+MODEL = "gpt-4o-mini"
+db_name = "vector_db"
+
+# Load environment variables in a file called .env
+
+load_dotenv(override=True)
+os.environ['OPENAI_API_KEY'] = os.getenv('OPENAI_API_KEY', 'your-key-if-not-using-env')
+
+def process_pdf(pdf_file):
+    loader = PyPDFLoader(pdf_file.name)
+    pages = loader.load()
+    
+    text_splitter = RecursiveCharacterTextSplitter(
+        chunk_size=500,
+        chunk_overlap=50
+    )
+    chunks = text_splitter.split_documents(pages)
+    embeddings = OpenAIEmbeddings()
+    #print(chunks)
+    if os.path.exists(db_name):
+        Chroma(persist_directory=db_name, embedding_function=embeddings).delete_collection()
+
+    # Embed the chunks with OpenAI Embeddings
+   
+    vectorstore = Chroma.from_documents(documents=chunks, embedding=embeddings, persist_directory=db_name)
+    
+    # Sample embedding dimension
+    collection = vectorstore._collection
+    sample_embedding = collection.get(limit=1, include=["embeddings"])["embeddings"][0]
+    dimensions = len(sample_embedding)
+    print(f"The vectors have {dimensions:,} dimensions")
+    
+    # Create the OpenAI Chat Model
+    llm = ChatOpenAI(temperature=0.7, model=MODEL)  # Or another model
+    response = llm.invoke([HumanMessage(content="Hello, how are you?")])
+    print(response.content)
+    # Set up conversation memory
+    memory = ConversationBufferMemory(memory_key='chat_history', return_messages=True)
+    
+    # Set up the retriever (vector store)
+    retriever = vectorstore.as_retriever()
+    
+    # Set up the Conversational Retrieval Chain
+    conversation_chain = ConversationalRetrievalChain.from_llm(llm=llm, retriever=retriever, memory=memory)
+    
+    # Return the conversation chain
+
+    return conversation_chain
+
+# Function to upload PDF
+def upload_pdf(file):
+    global chain
+    if file is None:
+        chain = None
+        return "pleae upload the file!"
+    chain = process_pdf(file)
+    return "processed the file ask questions"
+
+# ask_question function
+def ask_question(question):
+   
+    # Assume chain_holder["chain"] is your retrieval QA chain (e.g., ConversationalRetrievalChain)
+    if chain is None:
+        return "upload the pdf first"
+    
+    # Run the chain to get an answer
+    response = chain.invoke({"question": question})
+    
+    # You might need to adjust depending on what chain.invoke returns
+    if isinstance(response, dict) and "answer" in response:
+        return response["answer"]
+    else:
+        return str(response)  # Just return raw output if not nicely formatted
+
+# Building Gradio Interface
+with gr.Blocks() as demo:
+    gr.Markdown("## Chat with your pdf!!")
+    
+    # File uploader
+    file_input = gr.File(label="Upload your PDF 📄", file_types=[".pdf"])
+    
+    # Status text
+    status = gr.Textbox(label="Status", interactive=False)
+    question = gr.Textbox(label="Ask a question about the PDF")
+    answer = gr.Textbox(label="Answer")
+    ask_button = gr.Button("Ask")
+
+    file_input.change(upload_pdf, inputs=[file_input], outputs=[status])
+    ask_button.click(ask_question, inputs=[question], outputs=[answer])
+    
+    chain = None  # global QA chain
+
+# Launch the app
+demo.launch(inline=False)